Image segmentation using PCNN and maximal correlative criterion

After gray image being processed by special pre-processing method, using the maximal correlative criterion as the rule of iteration of PCNN (Pulse Coupled Neural Network), the gray image is segmented successfully. The information of image edge and fuzzy details is enhanced by special pre-processing with the mode of block shape enhancement, and the influences which different parameters of PCNN may come into being different image segmentation results are synchronously weakened. The computer simulation shows that better gray image segmentation results can be obtained using the characteristic of similar group pulses firing of PCNN and the maximal correlative criterion as the rule of iteration of PCNN. Compared with other correlative references, our method shows more image details and has preferable practicability.